SpaceMoE: Towards Orbital General Intelligence with Distributed Mixture-of-Experts Inference

TL;DR

SpaceMoE proposes a distributed mixture-of-experts framework tailored for satellite networks to enable scalable, efficient, and sustainable on-orbit inference of large language models amidst satellite-specific constraints.

Contribution

It introduces the SpaceMoE paradigm, addressing expert placement, selection, and routing challenges specific to satellite networks for on-orbit AI inference.

Findings

Reviewed industrial progress and standardization trends in space AGI.

Outlined fundamental design problems and solutions for SpaceMoE.

Highlighted satellite-specific factors affecting MoE architecture.

Abstract

As satellite networks evolve to support increasingly diverse services and artificial general intelligence (AGI), large language models (LLMs) are emerging as a critical foundation for future space systems. However, deploying LLMs on satellites is hindered by stringent constraints on onboard memory, computation, and energy. In this context, the mixture-of-experts (MoE) architecture emerges as a promising solution, leveraging sparse expert activation to enable scalable model inference. By harnessing the architectural advantages of MoE, this article provides a comprehensive overview of SpaceMoE, a new paradigm for distributed MoE inference in satellite networks. We first review recent industrial progress and emerging standardization trends that motivate the evolution toward space AGI systems. Then, we introduce the fundamentals and architectural evolution of SpaceMoE. Subsequently, we discuss three fundamental design problems in SpaceMoE, namely expert placement, expert selection, and hidden-state transmission and routing, highlighting how satellite-specific factors such as dynamic topology, battery degradation, and thermal limits fundamentally reshape their solutions. Finally, we outline promising research directions for realizing scalable, efficient, and sustainable on-orbit MoE inference in future satellite networks.